Emulator-enhanced extreme event attribution for data scarce developing countries

Attribution of extreme events in developing countries poses a significant challenge. A primary hindrance is the lack of historical observations, which not only limits the appraisal of the extent of an extreme event, but also restricts benchmarking of climate models for the region. A secondary hindrance is that tropical climates, characteristic of developing countries, contain large uncertainties due to natural climate variability, which many climate models struggle to represent. As it is those countries and world regions where some of the most severe consequences of climate impacts emerge, addressing these challenges to robust climate attribution is critical to improve prospects of climate litigation in developing countries. In this study, we present a novel method for attribution using the Earth System Model (ESM) emulator for spatially resolved monthly temperatures, MESMER-M (Nath et al. 2022). We use a bootstrap method in calibrating MESMER-M, so as to also characterize its intrinsic parametric uncertainty. Attribution using MESMER-M is then demonstrated on the prolonged heat conditions of March/April 2022 over the Indo-Pakistani region. The outcomes of this study are twofold. Firstly, by calibrating MESMER-M on the BEST observational dataset, we are able to inflate observational records with observationally consistent natural climate variability estimates, enabling exploration of “possible pasts” and insofar characterization of the event and its likelihood under rising Global Mean Temperatures (GMTs). Secondly, by exploring the parametric uncertainty space of MESMER-M calibrated on both BEST and ESM data, we systematically disentangle the uncertainty surrounding the mean response of monthly temperatures to GMT from that surrounding the natural climate variability. Such allows robust appraisal of the uncertainty surrounding natural climate variability as present within ESMs/Observations for the region, so as to not over/understate the event’s likelihood under rising GMTs.

 

Nath, S., Lejeune, Q., Beusch, L., Seneviratne, S. I., & Schleussner, C. F. (2022) MESMER-M: an Earth system model emulator for spatially resolved monthly temperature. Earth System Dynamics, 13 (2), 851–877. doi: 10.5194/esd-13-851-2022